LANDSLIDE SUSCEPTIBILITY ASSESSMENT OF KOKAP AREA USING MULTIPLE LOGISTIC REGRESSION

A number of landslides have occured in Kokap SubDistrict, Kulon Progo Regency, Yogyakarta Special Province, Indonesia, which have influenced the communities. The natural disaster is commonly associated with a few days of heavy rainfall events. To mitigate the impact of landslides in this area, a landslide susceptibility assessment needs to be carried out. The main objective of this research is to develop a landslide susceptibility zonation in the research area by applying a logistic regression (LR) method. Field observation was conducted at 68 locations in the research area, in which 46 landslides occured. Data of slope angle, lithology, geologic structure and groundwater conditions were collected. The relationship between landslide occurrence and the slope angle, lithology, geologic structure and groundwater conditions was analysed using the LR method. The analysis results showed a 0.984 standard error, implying a good-fit model. The study area was classified into very low, low, moderate, high and very high landslide susceptibility zones with 0–20%, 20–40%, 40–60%, 60–80%, and 80–100%, respectively, probabilities of occurrence. A 60% area of the total study area was classified as a moderate to very high susceptibility to landslide. From 47 landslides, 80% landslides occured in high and very high landslide susceptibility zones, 17% landslides occured in the moderate susceptibility zone and 2% landslides occured in the low susceptible zone. None of landslides occured in the very low landslide susceptibility zone. The analysis results show that LR method is a very useful method for landslide prediction.

A number of landslides have occured in Kokap SubDistrict, Kulon Progo Regency, Yogyakarta Special Province, Indonesia, which have influenced the communities. The natural disaster is commonly associated with a few days of heavy rainfall events. To mitigate the impact of landslides in this area, a landslide susceptibility assessment needs to be carried out. The main objective of this research is to develop a landslide susceptibility zonation in the research area by applying a logistic regression (LR) method. Field observation was conducted at 68 locations in the research area, in which 46 landslides occured. Data of slope angle, lithology, geologic structure and groundwater conditions were collected. The relationship between landslide occurrence and the slope angle, lithology, geologic structure and groundwater conditions was analysed using the LR method. The analysis results showed a 0.984 standard error, implying a good-fit model. The study area was classified into very low, low, moderate, high and very high landslide susceptibility zones with 0–20%, 20–40%, 40–60%, 60–80%, and 80–100%, respectively, probabilities of occurrence. A 60% area of the total study area was classified as a moderate to very high susceptibility to landslide. From 47 landslides, 80% landslides occured in high and very high landslide susceptibility zones, 17% landslides occured in the moderate susceptibility zone and 2% landslides occured in the low susceptible zone. None of landslides occured in the very low landslide susceptibility zone. The analysis results show that LR method is a very useful method for landslide prediction.